Utilizing the GPU with c#

Question

I am trying to get more processing power out of my grid.

I am using all cpus/cores, is it possible to utilize the GPU with C#.

Anyone know any libraries or got any sample code?

Answer 1

[Edit OCT 2017 as even this answer gets quite old]

Most of these answers are quite old, so I thought I'd give an updated summary of where I think each project is:

• GPU.Net (TidePowerd) - I tried this 6 months ago or so, and did get it working though it took a little bit of work. Converts C# kernel code to cuda at compile time. Unfortunately their website has been down and their github hasn't been updated for a couple of years, which might indicate the project is dead....

• Cudafy - Open source and very easy to use. Converts C# kernel code to cuda at runtime (with ability to serialize and cache). Can easily run the same kernel code on the CPU (mostly for debugging). Supports multiple GPUs. More examples available than others here. The boilerplate code referred to by other answers is minimal, and in my case at least helped with my understanding of how the code works. Cuda/Nvidia only though. Unfortunately, it seems that they didn't update their solutions for a couple of years too (latest commit in 2015 -- support of cuda 7.0).

• Hybridizer. Commercial solution compiling C# to CUDA. Provides a free community edition on visual studio marketplace and samples on github.

• AleaGPU Commercial solution with a free community edition for consumer GPUS. See Daniel's comments for details.

• Brahma - runs LINQ expressions via OpenCL (so supports AMD too). Not much documentation / examples. Last update in 2011.

• C$ - last development was over 10 years ago...

• Microsoft Accelerator - similarly doesn't look like it is being actively developed any longer.

• some others (C++ AMP, OpenTK -- dead/Cloo) - many of these are just bindings - ie enable you to call the GPU from C#, but your kernel code (code which is actually run on the GPU) needs to be written in C or OpenCL, meaning you must use (and learn) another language.

As I said, I would recommend Cudafy over all the others - if it could run on OpenCL as well as Cuda it would be perfect.

EDIT SEP 2013 Cudafy now allows you to compile for both CUDA and OpenCL, so will run the same C# code on on all GPUs. This sounds fantastic, though I haven't tested the OpenCL compiling yet.

Answer 2

Microsoft Research Accelerator was a .NET GP GPU library.

Answer 3

I found Brahma... It also has a GPGPU provider that allows methods to run on the GPU... Thanks for the question... Learnt something new today. :)

Answer 4

Here's another one: CUDAfy. It sounds like GPU.Net, in that something as simple as a method-attribute can cause the entire method to run on the GPU. But unlike GPU.Net, CUDAfy is free and open-source.

GPU.Net appears to require no boilerplate code, though (According to their docs, it's "injected automatically by the build-tool"), while CUDAfy does.

---

Here is an example of building an application with CUDAfy.

Answer 5

Could I recommend XNA Game Studio as a possible avenue for exploration? It is obviously geared up for writing games, but gives you managed access to your graphics card and much better access to capability enumeration functions and shader development than was previously available in, say, Managed DirectX. There are also ways of combining WinForms and XNA into hybrid applications:

http://www.ziggyware.com/news.php?readmore=866

You'll have to put some effort into learning shader programming (XNA supports HLSL), but this may be a simpler approach than learning a vendor-specific solution such as nVidia's CUDA. The advantage is that you can program in a 100% managed environment. Here are some HLSL links:

http://www.ziggyware.com/weblinks.php?cat_id=9

The GPGPU site is also a recommended destination for general purpose GPU programming:

http://gpgpu.org/

Best of luck!

Answer 6

How about http://www.tidepowerd.com/ GPU.NET ?

Answer 7

Well this is a pretty old question, and since it's been asked things have changed a lot.
Another option for using .Net to write GPU code, which no one has mentioned in answers in Alea GPU. It covers C#, F# and VB.

Professional GPU software development environment for .NET and Mono. Truly cross-platform

In F# official site, Alea is the first option for using F# in GPGPU programming.
To get to know this framework I suggest take a look at its comprehensive list of examples.

Answer 8

There's a new Microsoft solution in town - C++ AMP (intro here).

Use from C# would be via P/Invoke, as demoed here for desktop apps, and here for (don't-call-it) Metro apps.

Edit: I should note that C++ AMP has an open specification, which means it's not necessarily just for the MS compiler, or just for Windows.

Edit: Apparently, the technology is now in "maintenance mode," meaning they're fixing bugs, but not actively developing.

Answer 9

In addition to Brahma, take a look at C$ (pronounced "C Bucks"). From their CodePlex site:

The aim of [C$] is creating a unified language and system for seamless parallel programming on modern GPU's and CPU's.

It's based on C#, evaluated lazily, and targets multiple accelerator models:

Currently the list of intended architectures includes GPU, Multi-core CPU, Multi-GPU (SLI, CrossFire), and Multi-GPU + Multi-CPU Hybrid Architecture.

Answer 10

CenterSpace Software has GPU-powered computation in their NMath libraries you can add to C# project. It's a commercial product.

Answer 11

If your GPUs are all the same brand, you might be able to get GPGPU support from the vendor, either through Nvidia's CUDA or ATI's Stream. AFAIK, they provide DLLs, which you could use through P/Invoke.

Answer 12

Managed DirectX somehow, might work

Answer 13

WPF also uses the GPU and you can add custom shaders using HLSL.

• A Series on GPU-based Effects for WPF (Greg Schechter's Blog)